Re: [dpdk-dev] [PATCH v4 2/2] event/sw: support device stop flush callback

[Jerin Jacob <jerin.jacob@caviumnetworks.com>]

-----Original Message-----
> Date: Mon, 2 Apr 2018 15:50:42 +0000
> From: "Eads, Gage" <gage.eads@intel.com>
> To: Jerin Jacob <jerin.jacob@caviumnetworks.com>
> CC: "Van Haaren, Harry" <harry.van.haaren@intel.com>, "dev@dpdk.org"
>  <dev@dpdk.org>, "hemant.agrawal@nxp.com" <hemant.agrawal@nxp.com>,
>  "Richardson, Bruce" <bruce.richardson@intel.com>,
>  "santosh.shukla@caviumnetworks.com" <santosh.shukla@caviumnetworks.com>,
>  "nipun.gupta@nxp.com" <nipun.gupta@nxp.com>
> Subject: RE: [PATCH v4 2/2] event/sw: support device stop flush callback
> 
> If it's ok with you, I'll resubmit these two patches separately. The sw implementation has some unresolved race conditions with the scheduler service that may take some time to fix, but this shouldn't block the first patch. That patch needs a fix too (v4 includes a now-incorrect comment about stop not draining events from ports), so I'll fix and resubmit that.

OK. So that we can merge this common code and octeontx changes.

How about changing the following comment rte_event_dev_stop()

exiting:

+ * This function does not drain events from event ports; the
application is responsible for flushing events from all ports before stopping the
device.

proposed:(Just to share the view)
# You can add the fact that, all events across the ports will show up in
callback as we concluded.

#Application may invoke stop operation on the event adapter to
avoid event being generated.(or something similar) i.e adapter needs to
stopped before stopping the eventdev to avoid continuous looping on
getting the events(over rte_event_dev_register_callback()) from event source(ethdev Rx adapter).


> 
> Thanks,
> Gage
> 
> > -----Original Message-----
> > From: Jerin Jacob [mailto:jerin.jacob@caviumnetworks.com]
> > Sent: Monday, April 2, 2018 3:03 AM
> > To: Eads, Gage <gage.eads@intel.com>
> > Cc: Van Haaren, Harry <harry.van.haaren@intel.com>; dev@dpdk.org;
> > hemant.agrawal@nxp.com; Richardson, Bruce <bruce.richardson@intel.com>;
> > santosh.shukla@caviumnetworks.com; nipun.gupta@nxp.com
> > Subject: Re: [PATCH v4 2/2] event/sw: support device stop flush callback
> > 
> > > > [Update: Ah I see you do this below already. Still, better check
> > > > twice I think, the data path isn't running here anyway in case
> > > > future me decides to call
> > > > sw_flush_iq() without performing the check]
> > >
> > > Agreed, will fix in v5.
> > 
> > Gage,
> > 
> > Could please send the v5 so that I can include it in RC1 pull request.
> > 
>